Merissa Piazza
Welcome. I’m Merissa Piazza from the Federal Reserve Bank of Cleveland. Welcome to day four, the final day of the Uneven Outcomes in the Labor Market Conference. This conference was organized by the community development staff from the Federal Reserve Board and the Federal Reserve Banks of Atlanta, Boston, Cleveland, Philadelphia, San Francisco, and St. Louis. We designed this conference to convene researchers, policymakers, and practitioners to examine disparities in labor market outcomes and explore policy solutions to address these inequities.
We aim to deepen your understanding of disparities employment, labor market participation, income and wealth, and to learn about their implications for economic growth, the health of communities, and individual wellbeing. We welcome questions throughout the session. Please enter any questions into the Q&A function. The chat function has been disabled for the session. We titled this panel today, What’s Government Got To Do With It? Policy At All Levels. We will be discussing how federal, state, and local policies could be used to close uneven outcomes in the workforce.
Full bios for all speakers are available at the conference website. A link is available in the Q&A box on the screen. Today, you’ll hear from the following speakers. Bill Rodgers, Vice President and Director of the Institute for Economic Equity at the Federal Reserve Bank of St. Louis, he will provide the framing for today’s topic. Raji Chakrabarti, Head of Equitable Growth Studies at Federal Reserve Bank of New York. She will present her research on the effects of changes in the federal funds rate on households. Kathryn Rudloff, Executive Director of weVENTURE Women’s Business Center at the Florida Institute of Technology.
She will present her research on the impact of state-level family policies on female labor force participation and entrepreneurship. Heather Stephens has also joined us. She’s Associate Professor and Director of Regional Research Institute at West Virginia University. She will present her research on how economic development incentives affect racial and gender segregation of employment in regions. Neil Ericsson, principal economist at the Federal Reserve Board. He will discuss his research on labor force participation and unemployment structural change from the pandemic.
Finally, Enrique Lopezlira, Director of the Low-Wage Work Program at the University of California Berkeley. He will discuss the research and moderate the question and answers. During today’s session, the views expressed are those of the presenters and not necessarily those of the Federal Reserve System or the Board of Governors of the Federal Reserve System. I’m so excited to have Bill Rodgers talk to us today.
Bill Rodgers
Hi, Merissa, and thank you so much for leading us off. It’s truly an honor and a privilege to have the opportunity to open up what’ll be an amazing panel. I promise you. As my task was to really provide about 10 minutes of just framing remarks around this topic. The first thing I want to do is talk about or identify what are some uneven outcomes in terms of how do we measure that as economists or social scientists? What are some economic indicators? The ones that we focus on at the Institute for Economic Equity are employment population ratios, which capture not only search, via the unemployment rate, but also capture decisions to participate.
That rolls up into the employment population ratio. That’s one indicator that we’re very important. We’re very concerned about it. Another, obviously an outcome is our hourly wages and hours worked. A third bottle focuses on what’s the intensity of your connection to the labor market and that of your working and are you working full time? Then, some of our papers will go beyond those basic measures and talk about wealth and debt. Then, the final area that we focus on at the institute with regards to key economic indicators is a relationship.
It’s a relationship between where an individual lives and the area’s economic health as proxied or measured by the area unemployment rate or proxied or measured by the job openings rate. What we have been finding there and in publishing some of our blog work is showing that another source of unequal outcomes, not only the levels, like lower employment population ratio, higher unemployment rates, lower wages, fewer hours work, but that when area economic conditions improve, we see a disparate positive impact for certain groups. Or when the job openings slow down, we see a disparate negative impact on groups, certain groups.
That leads us to, well, who are some of the folks that experience uneven outcomes? Well, through our work and other people’s work, we found that at the top of that list there are young non-college-educated individuals who have no more than a high school degree and they’re not enrolled. They truly are at an age between 16 and 24, but it’s really about in their eyes work or not work. A second group that consistently experiences uneven outcomes are adults with no more than a high school degree. These are people who are 25 and older who have no more than a high school degree.
They have either lower wages, lower employment population ratios. They tend to be more sensitive to changes in the macro economy. Adult women of all races or another group who consistently have uneven outcomes. Adult Black men and women, adult Latino men and women. Also, people with a disability. We’ve been showing in our work and others too have shown that there are these uneven outcomes for them. Then finally the other group, another group that we’ve identified are Native Americans or American Indian and Alaska Natives, that have a number of these outcomes that are below overall outcomes but also overall relative to some peer groups.
Well, the question then becomes for us, well why do we need to care about these uneven outcomes? Well, here’s some reasons why. There are many others and there are other vulnerable groups or groups that I didn’t touch on, but that we highlight in the discussion. But in terms of why we need to care about these uneven outcomes, well one is that these groups have major more economic vulnerability. Some of you may have heard of a concept called ALICE, where ALICE stands for Asset Limited, Income Constrained, Employed. It was constructed by the United Way of Northern New Jersey.
Basically, their estimate of ALICE, these are people who do not have enough income or resources to make ends meet, to have a survival budget. Today or the most recent estimates they have is about 41% of American households or US households are ALICE. That number has not gone down, even as we came out of the pandemic. A second major reason why we need to care about these uneven outcomes and think about policy and other approaches is persistence. That when you examine or study the Black-white wage gap, the gender pay gap, we see that Richard Freeman and I and others have done work, but we’ve seen that tight labor markets don’t fully erode many of these uneven outcomes.
Third reason why we should care, well, these gaps or these persistent gaps are inconsistent with the mandate of pursuing stable prices and maximum employment. Fourth, macroeconomic growth and conditions have disparate impacts that I’ve described. Then fifth, if we place this conversation in a social determinants of the health model, these uneven economic outcomes translate into health outcomes. Then finally, that these uneven outcomes place a drag on productivity growth and ultimately economic growth. Well, this will be my last set of comments that will really jettison into the paper presentations on various policies.
That is the unfortunate piece when we start talking about policies, and this is why we have this panel here of multiple papers. There’s no silver bullet solution to addressing these uneven outcomes. The way I think about this work is three buckets, labor supply, labor demand, and institutions, and our authors are going to be presenting the work in these different areas. Particularly, labor supply, the role of child and elderly care with regards to generating uneven outcomes or labor demand. What’s the role self-employment opportunities and entrepreneurship can play with regards to creating more even outcomes?
What role do economic development incentives provide with regards to, again, smoothing out these uneven contours of experiences and then into what I call the institutional piece. That’s where a lot of the public policy conversation centers, where first it’s on education and childcare services for children age three and under, workplace policies at the local, state, and federal level, what are the roles? We’ll talk about the role of economic transfers and financial benefits. Then, we’ll also talk about, I’ll be handing it off to my colleague from the New York Fed in a minute who will talk about the real policies and their impact on the speed of the economy’s growth and that hints at their ability to create more even outcomes.
Then finally, what I think will permeate through all of these papers in a lot of our discussion is the role of societal norms, culture and history, such as the various policies around redlining that are decades old, but they still have persistent effects today. With that, let me stop and I’ll turn it over to my colleague from the New York Fed, Raji Chakrabarti and she’s going to talk about federal funds rates and policy associated with their movement, so thank you.
Rajashri Chakrabarti
Thank you very much, Bill. Thank you for having me present this paper today. This is joint work with Maxim Pinkovskiy, also at the New York Fed. I should say that it is a special honor to really follow Bill here just because Bill has done quite a bit of work, I should say seminal work in this area. Long time back, he was one of the first people to actually look at this and then really kind of say that there are uneven effects when you look at the labor market, a little bit of which he talked about today. It is really a pleasure to follow him and to hear him speak.
Of course, I’m very much looking forward to input at the end of this panel. With that, let’s go on to the next slide. In this paper, we are going to look at the effects of changes in federal funds rate on households. We are going to look at basically two channels. One is through the household debt market and the other one is the labor market. We are also going to ask whether the effects are different across different types of households. Is it different in the household market, as well as the household debt market, as well as in the labor market?
We start with different measures of exposure to the federal funds rate, and these would be different in the household debt market versus the labor market. Our idea is to understand whether households in areas with higher financial strain are affected differently and whether or not households in different markets, labor markets that vary by the tightness of the labor market are also affected differently. In turn, we’re going to ask, and the crux of this paper and presentation would be to understand whether these effects in turn vary by demographics and geography.
Here in this paper today, I’m going to focus on demographics. Let’s move on to the next slide. This is just the plot of the federal funds rate from ’99 until the present. This is the time span of our data set and our paper. As you can see, there has been lots of ups and downs depending on the time period under consideration. There is quite a bit of variation in the federal funds rate during our period of consideration, and that’s what we are going to leverage in this paper today. Let’s move on to the next slide. As I said, apart from the changes in the federal funds rate, the time series that you just saw, we are going to try to distinguish between areas that vary in terms of financial stress, which will be our first measure.
Then, we will also distinguish between markets that vary in terms of labor market tightness. In terms of the first measure, we use debt-to-income ratios. The idea here is for example, with FFR with federal funds rate increases, households with higher debt-to-income ratios or households that are more stressed financially may find it harder to originate new debt, it can be mortgaged debt or non-mortgaged debt, and make payments on existing debt, and consequently, incur more delinquencies. The effects may actually vary by the type of mortgages that the households hold.
For example, the effects for variable rate mortgages where your mortgage rate might change is quite different, would be quite different than if someone holds a fixed rate mortgage rate, in which case the effects of the FFR would be much less and much more muted. We in turn try to understand whether these impacts vary by demographics. We consider changes by differences by race, ethnicity, by income, by education, and by age, in this paper, or in today’s presentation. On the labor market side, we use employment-to-population ratios or vacancy to unemployment ratios to capture the tightness of the labor market.
The idea here is that with FFR increases, demand for workers may actually fall. If this does happen, then workers in a more slack labor market where you actually have a lot of this is basically a supply market, at that point of time, there may be adverse effects compared to in a tighter labor market. This again might vary by demographics, by race, ethnicity, by income, by education and by age. Let’s move on to the next slide where we are going to talk very briefly about the data we are going to use. We use a variety of data. Today, and we are going to focus mostly on the New York Fed/Equifax consumer credit panel.
Apart from that, we use the HMDA data, the McDash mortgage servicing data. We use pending data as well from Commerce Signals. We use House Price’s data and the labor market data that I’ll present today will come from the QWI or the quarterly workforce indicators. Let’s move on to the results. I’m not going to go into the methods. Let’s go to the next slide where the results are. I’m not going to talk about the methods, but just a one-liner. We use a local projection model with distributed lags and the results that you see here are plots from that specific model.
In this slide, we’re looking at the effects on FFR increases on the auto debt market and the first variable that we look at, at the impact on shares, auto shares, then auto balances, and then auto delinquencies as well. What we can find here, if we look at, for example, the first two charts, look at the first two charts. I should say, well, for every chart that I show you here today, there will be a blue line. The blue line will correspond to the overall effect, the total effect, as we call it, when in a geography that has a slightly higher financial stress or slightly higher DTI or debt-to-income ratios.
Then the red line will capture the effects in higher share bad Black areas where again, there is a slightly higher financial stress. Equivalently, the green line will capture the effects in higher share Hispanic areas with slightly higher financial stress. Now in the first two charts, what we can see that if FFR increases, at least initially, in the, let’s say the first five or six quarters, are associated with lower auto loan originations. Basically, lower car buying, you can think of that that way. Lower balances, lower auto balances as well.
That, we can see. This effect is actually much more pronounced for residents in higher share Black areas who are also residing in areas under higher financial stress. The last chart looks at the impact on auto delinquencies. Something that comes out very clearly in this chart is if you look at the effects after two and a half years, after 10 quarters, you see that the higher share Black areas that are also under higher financial stress pays a considerably higher delinquency rate than the other areas, than either the overall the average area or residents in higher share Hispanic areas.
Let’s move on to the next slide where we are going to look at the impact on mortgage debt. Again, a similar way of looking at the results, effects of FFR increases and whether or not the effects are different in areas under higher financial distress. Once again, are those different by the different geographies that have higher share of Blacks or higher share of Hispanics? What we can see here is if you look at mortgage delinquencies, the last picture, you can see that FFR increases are associated with higher mortgage delinquencies in areas under higher financial distress.
That comes out pretty clearly in that picture. But something that is possibly more clear if you continue to focus your attention on that same, that third chart, you can see that higher share Black areas that are also under higher financial distress see considerably higher delinquencies even when the delinquencies have started to decline in other places. That’s actually statistically different from those that we see in the other places. If we really quickly look at the first two pictures, in the first picture, we see an overall decline in mortgage originations with higher FFRs in more financially distressed areas, but they’re really not different across different demographic groups.
The middle chart, we again see there is lower balances as well in pretty much every area, not statistically different from each other. What really is prominent in this whole slide is a considerably higher delinquency of residents in higher share Black areas about two and a half years or so after the FFR increases. Let’s go on to the next slide. We can see these effects are pretty dynamic and some of these take time to actually show up, as you saw in the last slide. Now, this one is looking at the effect on credit card debt. The first one is on credit card limit, credit card balance and credit card delinquencies.
What we see here is that actually FFR increases are associated with higher credit card limits, there’s a typo here, in the first year and higher credit card balances and delinquencies in financially more stressed areas. Now, borrowers in higher share Black areas are prominently different in pretty much each of these pictures. You can see that borrowers in higher share Black areas that are financially more stressed from the first chart face perceptibly lower credit limits. That, you can see very clearly after about a year after the FFR shocks happen.
If you look at the last chart, you can see they also face higher credit card delinquencies and they continue to persist until the end of a time period, which is four years after the FFR changes, FFR increases. I should also point out that these differences are statistically different from each other. When I say that they’re statistically, the effects in the higher share Black areas are statistically different from that of the overall or the average effect. Credit card balances are also higher, you see from the middle picture, and that can very well be because of higher delinquencies in higher share Black areas.
One of the reasons we think that this might actually be behind this is that it can be higher credit card interest rates that follow higher FFRs. There may be credit market frictions that lead to lending aversions, for example, for people living in higher share Black areas, which may explain the lower credit limit and hence lower credit access, et cetera, which might actually lead to the patterns that you see in this slide. Let’s move on to the next one. This one looks at overall financial wellbeing and the first chart is looking at impacts on credit score.
The middle one is in foreclosures, and then, the last one is in bankruptcy. We actually see credit score declines as you can see, and bankruptcy increases, foreclosure increases as well after an FFR increase in areas under higher financial distress. Now, what we further see is that this increases or decreases. In the case of credit score, these are decreases. For foreclosures and bankruptcies, this increases. These are actually much more prominent in higher share Black areas that are under financial distress. These three measures are kind of summary measures that you can look at if we are trying to understand the impact on overall financial wellbeing.
These three charts clearly show that people residing in higher share Black areas that are under higher financial distress are actually relatively worse off after FFR increases. Some of these do take some time, but that is pretty clear here and these effects are actually different statistically from the overall or the average effect. Let’s move on to the next slide. This is going to be my final slide. Actually, there is one more quick slide with the charts. This basically tries to understand whether or not the effects that you just saw are contributed by differential labor market impacts also happening in the same areas.
We do see counties with higher financial distress, see both hires and separations decline, though hires decline more than separations do. As a result of that, there is a temporary drop in the employment rate in the first year after an FFR increase in more financially distressed areas. This pattern is higher in areas with higher Black share, but I should say that these are not statistically different from the overall effect, and magnitude wise, are small enough that they cannot explain the effects in the household debt market that you just saw. Final chart slide. Let’s move on to the next one.
In this next one, we just look at the labor market impacts of changes in federal funds rate and if they’re different in tighter labor markets versus not tight labor markets. Also, whether or not these are actually different across different demographic groups. The key finding here is that we see Black households, Hispanic households as well as female households, or workers, these are actually workers I should say, see greater declines in employment in the first five quarters after an FFR increase. This is relatively more in tighter labor markets, but the opposite is true for 45 to 54 middle-aged workers.
Final slide is the conclusion slide, and basically, to conclude here. In this paper, we try to understand the effects of changes in FFR on households. We look at both the household debt side as well as the labor market side. We distinguish between markets that vary by debt-to-income ratios, capturing higher or lower financial distress, as well as employment-to-population ratio that captures variation in the labor market tightness. We also look to see whether the impacts are different across different demographic groups. What we find here is that counties with higher debt-to-income ratios, that means countries that are under higher financial distress, actually have higher debt balances, higher delinquencies, higher foreclosures, lower credit scores and higher bankruptcies.
These impacts are in turn more pronounced in areas that are higher share Black areas. I should say that I didn’t point out during the presentation, but I should say we do not see this in areas that are higher shared Hispanic. Their effects are considerably more similar to the overall effect. Then finally, employment is more responsive to the changes in FFR in counties with high DTI, but they really cannot explain the total effect that we see, and in general, they are more responsive in areas that have higher labor market tightness in terms of employment. I’ll stop there and pass it on to Kathryn.
Kathryn Rudloff
Thank you so much and good afternoon. I’m grateful for the opportunity to share my ongoing dissertation research with you all today. All insights and analysis presented are my own and not reflective of Florida Tech University nor the Census Bureau or other organizations whose data is referenced. There has been a significant increase in the number of women-owned businesses in the United States over the last 30 years. In the five years leading up to the pandemic, women were starting an average of 1800 new businesses a day, far out-passing their male colleagues.
This trend continues during the economic recovery with women of color leading the way. This is a data point that organizations like mine often celebrate. But at the same time, the United States has also seen significant growth in the number of individuals, men and women classified as self-employed, which includes independent contractors. There’s also been an increase in the number of Americans working multiple jobs. While the growth of the gig economy has gotten a lot of attention, numerous analyses estimate the gig work task-based work facilitated by a web platform only accounts for between one and 5% of the labor force.
As we heard in one of the great presentations yesterday, research on self-employment is challenging, because self-reported survey data may be unreliable, as many individuals who fall into this class of worker often don’t have a clear understanding of themselves if they’re an employee, a contractor, or a small business. In 2017, the Census Bureau began producing a new data set, the Nonemployer Statistics Dataset, or NESD, to get a more complete picture of US business ownership by demographics. NESD leverages existing individual level administrative records to assign demographic characteristics to a universe of nonemployer businesses.
This new approach provides a much clearer picture of the class of worker I affectionately call solopreneurs. First slide or next slide. The first NESD data set released found that women-owned businesses made up 42% of all businesses in the US, over 13.1 million, but only one million of those had employees. This revelation is not unique to women. There were over 31 million small businesses with under 500 employees, but the vast majority of those 81% were also nonemployer firms. Next slide. Research following the structuralism approach suggests that lacking open and equal career opportunities, due to discrimination, marginalization, or other blocked opportunities pushes many racial, ethnic and gender minorities into entrepreneurship.
The most recent NESD data set seems to support this disadvantaged worker theory as businesses started by women and other racial and ethnic minorities are much more likely to be nonemployer firms. Recent research by LIM, in 2019, found the fastest growing segment of independent contractors were female heads of household in the bottom income quartile. Operating as a solopreneur rather as a primary or a secondary job suggests that individuals are not interested in building a company that creates traditional measures of economic productivity such as jobs and corporate tax revenue.
Rather, research has shown that women increasingly use self-employment to reduce conflicts between work and family considerations. Further, findings along both developed and developing economies in countries with high and low GDP indicate that women turn to entrepreneurship due to unequal opportunities to participate in the labor force. This type of work is described as a precarious form of employment, increasingly associated with lower wages, reduced earning potential, reduced pension and savings, and increased vulnerability to economic shock.
Solopreneurs do not benefit from worker protections, like anti-discrimination laws, OSHA collective bargaining, workers’ comp and unemployment insurance, nor were they included in the initial round of COVID relief. It took Congress six months to recognize that PPP had done nothing to help the over 25 million self-employed solopreneurs. Next slide. I wanted my research to explore this further. When I dove into the data, I found there was significant variation between the states and between the genders for the outcomes I was most interested in, labor force participation, self-employment and solopreneurship.
As my focus turned to the public policies that impact these outcomes, I recognized a gap in the literature. Comparative analysis within the United States is often narrowed to one policy or one program at a time. Few studies have taken a more comprehensive perspective of the variation of welfare state policies at the state level. Next slide please. That’s a problem, because federalism is fundamental to understanding the variation of outcomes. Most family policy comparative research is done at the national level, where too often it feels like the US is the control group with no national policy in place.
In the absence of federal leadership, many states have begun to experiment with policies and programs designed to support individuals with children so they can remain active in the labor force. Further, states have a significant influence on how federal funds are distributed. While composite indexes are frequently used in international comparative research to help summarize and communicate complex policy environments, domestically, they’re seen mostly in advocacy organizations and scorecards. I found only one recent academic paper that utilized a comprehensive cluster analysis approach to study policy variations and the impact on poverty. Next slide please.
This inspired my central research question and methodology. I’ve created a composite index of family supportive policies incorporating 28 indicators as identified from previous research. My four hypotheses are that states with more family supportive policy environments will see higher female labor force participation, will see a narrowing of the gender gap between men and women in the labor force. Since women in those states enjoy more worker protections and benefits, they’ll be more likely to stay in traditional wage labor. I hypothesize a decrease in self-employment and solopreneurship rates for women. Next slide please.
The 28 policy indicators are organized into three dimensions, childcare, economic transfers and benefits, and workplace protections. They are summed to provide a total index score. This slide lists the policy indicators and highlights the variation between the states. Notice, while all states do invest in early childhood education and care, there’s significant variation as to the generosity and enrollment rates. Looking in the middle column, notice the majority of states have implemented economic policies to support working parents.
However, there’s significant variation in the generosity in enrollment rates in these programs. This is in contrast to the final dimension, workplace protections, which finds very few states have implemented workplace policies to support working parents. I wish I had more time to discuss these policies in depth, but the main point I hope to make on this busy slide is that the whole process of creating a comprehensive index, it became obvious just how varied the state policy environments really are. Next slide please. This map shows the state rank for the total composite index score divided into thirds.
Vermont is the state that scored the highest overall and is ranked number one and Alabama scored the lowest and is ranked number 52. You can see the coastal states have implemented more policies. The Midwest, Virginia, the Carolinas, and Louisiana are in the mid-range and the majority of states in the Southwest, South and West are in the bottom third. My full analysis looks deeper into the rank for each of the three dimensions as well, but there’s not enough time to share that today. Next slide please. Looking at top line results from the correlation analysis, I found a significant positive relationship with female labor force participation rate and a significant negative correlation to the gender gap.
Meaning the first two hypotheses were supported, states with more family supportive policies do see strong female attachment to the labor force. The third hypothesis was not supported, as results indicate a positive but not significant relationship to the rate of self-employment. The results for solopreneurship were a bit mixed as the hypothesis was supported, showing a negative correlation, but not at a significant level. It’s also interesting that the workplace policy dimension, the one with the fewest states that currently have policies in place, is the only dimension that did not produce a significant impact. We have work to be done. Next slide please.
I also conducted regression analysis using an empirical model that incorporates state level and regional fixed effects. Again, the first two hypotheses were supported with significant effect and the R-square does indicate a good model fit across all variables. However, results did not support my hypothesis that a high index score would be associated with lower rates of female self-employment and solopreneurship. Some of the additional co-variates, like the state unemployment rate, had a very significant effect. Next slide please.
In conclusion, significant policy variations between the states have resulted in a variation of economic environments. My initial analysis suggests an interesting correlation with the uneven outcomes between the states for the three dependent variables considered. It is my hope that the creation of the US Family Policy Composite Index has made a significant contribution and will encourage more comprehensive comparative studies within the United States. Thank you. The next speaker is Heather Stephens.
Heather Stephens
Hi. I’m having trouble doing my video. Okay, thank you. If you could go back to the introduction slide please. Thank you for the opportunity to present this research. This is some joint work with my colleague Carlianne Patrick at Georgia State University. This is part of a series of papers that she and I have been working on and sort of in our overall research agendas, both of us looking at economic development incentives and their role in various labor market and other economic development outcomes. Today, we were talking about, or this week we’re talking about uneven labor market outcomes and disparities in employment and wages.
We really wanted to see, do these policies, economic development incentives, have the desired effect for all populations? Next slide please. Before we get going, I want to talk a little bit about what economic development incentives are. They are the primary labor market policies used by state and local governments to promote job opportunities. What I mean by that is targeted tax breaks, financial incentives, such as property tax abatements or job creation tax credits or grants. The goal of these policies is to increase employment in particular locations.
Previous research has shown that they can create jobs, although their results are a little bit mixed. Now as we’ve been discussing all week, we know that industry and occupational segregation by race and sex are pervasive features of US labor markets. However, little is really known about how this predominant policy to promote local economic growth could lead to wage or employment segregation by race or gender, and also, how it might sort of lead to income inequality overall. Next slide please. I’m going to first start to talk a little bit about the data that we’re going to use to define economic development incentives.
We have economic development incentive data from the WE Upjohn Institute Panel Data on Incentives. This data set covers 45 industries in 47 cities, which we are defining using the core-based statistical areas that are proxies for metropolitan areas in 33 states. This covers about 92% of the private sector GDP in 2013, so it covers a lot of the country, but we’ll admit that our data set doesn’t cover every place. We’re going to use the net taxes paid for each industry in each CBSA to reflect the incentives received. Because of that, when we get to the results in just a minute, you have to remember that higher net taxes means less incentivized or lower net taxes are industries with more incentives.
Then, we’re going to use these data to examine the relationship between these economic development incentives and industry segregation and wages. Now, it is possible that the industries that are being incentivized might already be segregated or might already be high-wage industries. We’re going to use a statistical method called instrumental variables regression to isolate the impact of incentives on employment and wages in these cities. To do so, we’re going to exploit national-level trends and employment and state constitutional provisions that date back to the 19th century, which allow for the use of incentives differently across states. Next slide please.
First of all, I’m going to just show some simple correlation charts to show you that there does seem to be some relationship between economic development incentives, industry segregation by race and gender, and differentials in wages. These two slides, sorry, this slide shows how economic development incentives, or net taxes, are correlated with the share of minorities in an industry and the share of female workers in an industry. If you look at the lower end of the net taxes, remember those are the more incentivized industries, we see more incentivized industries have lower shares of both minority and female workers.
Or conversely, higher net taxes, less incentivized ones are the ones that actually are more equal. Next slide please. Again, lower incentives, sorry, higher incentives is lower net taxes. Here we’re going to look at the differential in wages between men and women and non-Blacks and Blacks. You can see on the first chart here that lower incentives are associated with a, excuse me, lower non-Black wage premium. In other words, higher incentives basically means that white workers are making more. Similarly, on the right-hand side, you can see that higher incentives mean that males make a higher wage premium compared to female workers. Okay, next slide please.
All right, so now let’s turn to the results of our regressions. Because we have a lot of results, we are depicting these again using some figures. What our results find is that incentives do have a positive effect on wages. That’s good for policymakers that are interested in increasing employment and also increasing wages. But probably not surprisingly, we do see that there are differential impacts on wages where females and Black workers are actually benefiting less from this increase. Next slide please. We also see that the majority of this wage effect is for high wage workers.
High wage workers are seeing about a 2.4 to 2.8% increase in their hourly wages, while lower wage workers and middle wage workers are seeing either no or potentially a decrease in their hourly wages. This is obviously problematic, because it’s contributing to income inequality. Next slide please. Then we create a couple of measures that sort of look at the relative incentives that industries are seeing or getting. The first one, on the left-hand side, is our specialization index. This one is, is an industry more incentivized in a particular city compared to in other cities?
In other words, if it’s a manufacturing industry, is manufacturing in industry A more incentivized than manufacturing is in other cities? Then our targeting index says, within a city itself, is that particular industry getting more incentives? In both cases, we see some evidence that this is contributing to wage disparities where the male to female wage differential is higher in more incentivized industries. Similarly, the non-Black to Black. I guess I’ll just remind you that just like on the other slides in this case, a lower index is more incentivized in a more incentivized industry. Next slide please.
The most important thing here is that economic development incentives have been shown to have some pretty beneficial effects in some places in creating jobs, although the effects in other places are mixed. There’s also been some previous research that Carlianne Patrick and I have done that shows that it may be having some crowding out effects for middle wage workers. We see some of that evidence here. But the real big problem is that if we’re concerned about helping everyone or if we’re concerned about having even outcomes in our labor market, these results suggest that not everyone may be benefiting equally.
It seems to be that we are actually potentially leading to more inequality by these incentives, continuing segregation and continuing wage disparities by gender and race. This suggests that when we are thinking about sort of policy analysis and implementing new policies to support economic growth, we need to think not just about the big picture of whether this creates higher wages or more jobs, but also what are the differential impacts? Obviously, this whole session is about that, and so I’m really pleased that we were able to share these initial results. With that, I’m pleased to go ahead and turn this over to our next speaker, Neil Ericsson. If you want to reach me, there’s my information. Thank you very much and thank you again for this opportunity to present today.
Neil Ericsson
Thank You. I’m delighted to be here. Full thanks to the organizers, especially to Megan and Heidi. Usual disclaimer applies, and Bill provided an excellent motivation for what I’ll be talking about here, which is joint research with Victoria Tribone at Johns Hopkins and Andrew Martinez of the US Treasury. Bill emphasized that one of the key variables to look at was the employment-to-population ratio and we’ll be decomposing that, in effect, into the labor force participation rate and the unemployment rate. We’ll also mention the importance of persistence of disparities and will be focusing very much on that. Next slide please.
As a little background, I come at this from two quite different perspectives. One is that I am a Federal Reserve Board economist where I have a long-standing interest in both empirical modeling and economic forecasting and evaluating and improving those forecasts that are being used in policy decisions. I’m also a teacher in the American Economic Association summer program, which is currently at Howard University. The program aims to increase the diversity in the field of economics by preparing talented undergraduates for economics PhD programs and has been very successful.
Roughly 20% of economics PhDs awarded to minorities over the last couple of decades have gone to the summer program graduates. This is a micro-based attempt, if you wish, micro-based policy to try to even out some of the disparities that we see. Next slide. Where unequal outcomes are is actually central to implementing policy. As an example, Federal Open Market Committee policy is very highly data-based, so understanding the data is critical. Both former Chair Janet Yellen and the current Chair Jay Powell have emphasized this in any number of speeches that they’ve given including Jay Powell at its most recent post-EMC press conference.
The committee will carefully assess the incoming data. The Fed’s Monetary Policy Report to Congress also reflects the importance of the labor market in FOMC policy. Next slide please. Just under a year ago, here’s one of the highlights in the Monetary Policy Report. Why has the labor force recovery been so slow? Next slide. One of the graphs they give is of the labor force participation rate and it really shows how much the pandemic has influenced labor force participation. That’s going to be one of the key features or the key focuses of the current paper. Next slide.
Another concern is disparities across the labor force and the Monetary Policy Report has this graph of the employment rate by race and ethnicity. It’s an area that I’ve worked on with other people, but today, I’ll be focusing on gender and age. Next slide. The motivation and aims and framework of the presentation focus on documenting where these unequal outcomes are, looking at changes over time, whether changes are persistent or temporary. Also, especially, as we saw in the graph of labor force participation, the real distinction between pre-pandemic and post-pandemic, or pandemic and thereafter.
These disparities show up across subgroups and we’ll be looking at gender of age. We’ll be assessing these disparities in an innovative approach where we’ll use dynamic modeling of systems of economic variables, both of the labor force participation and unemployment rates broken down by subgroups and also across subgroups to pick up gaps in disparities. We’ll also forecast from a pre-pandemic model into the pandemic to understand the consequences in the labor market of the pandemic. To summarize, the marked disparities across gender and age, also across ethnicity, especially at the beginning of pandemic in 2020-21, lower labor force participation persists among many of the subgroups. Next slide please.
Details are that we’re going to focus on these two key labor force variables, labor force participation rate and US unemployment rate, is drawn from the sample of around 60,000 observations surveyed in each survey each time point. We’ll be looking at those fluctuations in unemployment rates and labor force participation rates across different sectors and the many potential implications, including labor shortages, economic recovery, hidden unemployment, and economic policy. Next slide. Two key variables, unemployment rate, which is the number of unemployed relative to the labor force, and then the labor force participation rate, which is the labor force relative to the population.
Bill mentioned the employment-to-population ratio, which is just a different way of representing those two variables together. Next slide. We’re going to begin by looking at the disaggregated by gender, both male and female activity in the labor force. You see a gradual narrowing of the gap in terms of labor force participation rates between men and women, pre-pandemic and also in the pandemic. But there’s persistently lower labor force participation rates in the pandemic. For unemployment rates, here there’s actually a typo, it’s relatively lower female unemployment rates during the financial crisis, but the relatively higher female unemployment rates early in the pandemic, which arose in fair part because of women tending to be more likely to be involved in the service industry.
The service industry was negatively affected by the pandemic. There’s been a rapid recovery in unemployment rates more recently. Next slide. The upper left graph gives us labor force participation rates for males, for females, and then the total. The lower left slide gives us the gap, the differential, and of course, we’re focused here on uneven outcomes, we can see that the gap is narrowing over the last four decades and even into the pandemic continued narrowing, but it’s still persistent. We’re still looking at a gap of 10 percentage points in the labor force participation rate.
The second column of graphs gives us the same service results for the unemployment rates, and again, seeing the notable spikes, both at the time of the natural crisis and then recently the pandemic. Next slide. We’re going to build a model over the pre-pandemic period and then use that to forecast the pandemic where we’re now going to interpret the forecast as representing what would’ve happened if the pandemic had not mattered for the labor model. Then use the gap between those forecasts and what actually occurred to pick up what the effects of the pandemic was on the labor force, on the labor market.
The disparate labor outcomes for the pandemic. The unemployment rates return to near-pandemic levels relatively rapidly, whereas labor force participation rates have been very slow to recover still below pre-pandemic levels and women were much more adversely affected by the pandemic. We’ll then go on and look at the disaggregated by both gender and age using the same sort of methodology. Next slide. This gives us forecasts of what would’ve happened if the pandemic hadn’t mattered compared to what actually happened. Looking at the upper left panel, this is the labor force participation rate for males, which would’ve continued more or less at its previous level of 69%, but in fact dropped to 67%.
Two percentage points is enormous force participation rates. Likewise for females, we’re seeing about a two percentage point drop in labor force participation rates, the lower left panel. Unemployment rates behave very, very differently. Instead of this very strong persistence, we see the spikes in the unemployment rate, but then within a year and a half of returning, nearly to the pre-pandemic levels. This also shows up in the disaggregated by race and ethnicity. Next panel. Here, we break it down not only by gender, but also by age.
The first column is 16 to 24 years olds, the second column 25 to 54 years old, third column is 55 and above. There are very, very different patterns that we see here. For instance, in the upper left panel, 16 to 24 year olds, we see a really quite rapidly narrowing gap between labor force participation rates in men and women, whereas we don’t see that sort of narrowing going on in older workers either in the middle panel or in the 55 and above. Next slide please. We then go and do the same exercise to forecast representing an as-if type of scenario.
This is looking at the 55 and above. The first column is labor force participation rate for males 55 and above, and we see this well-known phenomenon now. For the older workers, labor force participation rates still remain well below what they were pre-pandemic. We see that both in the first upper left panel for males and upper right panel for females. Whereas in the lower row, we see unemployment rates returning to very much the pre-pandemic levels and relatively quickly. Next slide. We see these uneven outcomes in the labor market. There are strong differences pre-pandemic and they’re also strong and somewhat different differences, if you wish, into the pandemic with unemployment rates returning near pre-pandemic levels by late 2021.
Women in all age groups feel much more adversely affected than then. Labor force participation rates continuing below pre-pandemic levels for many groups. The recoveries in labor force participation rates really do differ very markedly by age. For the youngest category 16 to 24 year olds, we see within a year, recovery to pre-pandemic levels, whereas 25 to 54 year olds, we see some recovery, and 55 and above, we see labor force participation rates really stagnating even up to the current period. This really sets the agenda for looking at these sorts of disparities across other sorts of factors, which you’ve already mentioned in the session by race, by job, by recreation. These are all very important things to be documenting, so we can set better policy. Thank you. I’ll turn it over to Enrique to discuss.
Enrique Lopezlira
Thank you, Neil. Good afternoon everyone. I get the pleasure of discussing these four great presentations and papers. Just given the time, I will not focus on methodology. For each of the authors, I will email separately some comments and thoughts on methodology and data issues related to your papers. I’m going to just focus my comments on the results and have some questions for the authors and then I’ll moderate any questions folks may have in the Q&A.
I’ll start with Raji’s paper, The Uneven Effects of Changes in Feds Funds Rate on Households, and so it’s an overview of her paper. She looks at the federal funds rate, which is the interest rate charged by banks to borrow reserves on an overnight basis. She’s exploring how changes to short-term interest rates may impact households. She looks at two different mechanisms, the effects under different market conditions, a strained financial market and a tight labor market. Then, she looks at whether these effects differ under those market conditions by demographic and geographic characteristics of households. For the financial market, she uses one measure, which is the debt-to-income ratio.
Then she looks at two labor market measures, vacancies to unemployment and EPOP, or the employment-to-population ratio. Then she, in her model, she looks at the impact of changes in the federal funds rate up to four years after the policy changes, either increases or decreases. Her results are that for areas on their financial strain, the FFR increases are associated with lower loan origination for autos, lower balances, the first one and a half years of auto loans and higher balances after and higher delinquencies over the first three years. Increase in the FFR is also associated with higher mortgage delinquencies, lower mortgage originations, lower mortgage balances.
The FFR increase is associated with lower credit card limits, higher credit card balances and higher credit card delinquencies in financially strained markets and in high labor markets. The FFR increase is associated with lower overall financial wellbeing. When looking at demographics and geographic characteristics, these effects are greater for higher share Black areas, especially in delinquencies. These differences are especially pronounced in credit cards, mortgage delinquencies and financial wellbeing. In tight labor markets, she finds that the FFR increase is associated with lower hires, lower quits and layoffs, and overall lower employment, and the labor market effects are more pronounced for women and minorities in the first year.
Those are the results and overview of her paper. I have a couple of comments. I enjoyed the presentation, and like I said, I will email separately on methodology. I find interesting that she looks at, even though the FFR is a short-term interest rate, she looks at a mix of short-term and long-term credit markets, right? She looks at autos and credit cards, which are short-term, and she looks at long-term in terms of mortgages. Obviously, short-term interest rates do impact mortgages, but I think mortgages, at least over the past few years, I’ve responded more to ten-year bond yields than fed fund rate changes.
I’m overall not surprised by the heterogeneity of effects by demographics, but I believe this speaks more to structural issues with financial unemployment markets more than FFR policy. In other words, increasing the price of funds has more pronounced effects on minorities and women, because they have more precarious financial market opportunities that lead to higher interest rate car loans, credit cards and mortgages, and so they’re more sensitive to interest rate changes. They’re also more likely to be in precarious labor market conditions, low-wage occupations with less stable employment and labor standards.
When funds get tighter, lenders begin tightening on more precarious loans, which like I said, minorities and women are more likely to be in. Similarly, there’s plenty of evidence that when the economy slows down, which is what the Fed tries to do with an increase in the federal funds rate, we see that if it slows down and unemployment goes up, Black workers tend to be the first ones out and the last ones in during the recovery. I see the FFR effects that Raji finds as symptoms of structural inequities in the two markets that she explores in the paper, the financial market and the labor market. For Raji, I would like to get her thoughts on comparison of FFR changes during this COVID crisis compared to the Great Recession.
I’ve seen some research that during the Great Recession business confidence dampened some of the FFR effects. For example, the lowering of the FFR rate during the Great Recession didn’t impact unemployment as much as expected, because the business confidence was low. Currently, we have FFR increases to deal with inflation, and we see that business confidence is again, particularly in small businesses, low. Any thought she has on comparing the two. Then, just overall reactions to how much of this is really FFR driven or FFR is just a symptom of some structural inequities that are baked into these markets that just magnify the impact of the FFR. Thank you, Raji.
I’ll move to paper two, which is Kathryn and The Impact of State-Level Family Policies on Female Labor Force Participation and Entrepreneurship. Just an overview, she explores how state-level family policies impact female labor force participation in the US. She focuses on self-employment and nonemployer firms or solopreneurship among women. I guess the question is, are precarious working conditions pushing women to self-employment or solopreneurship? As she mentioned, self-employment is not all that it’s cut out to be despite what many platforms say when recruiting folks to drive or deliver for their companies.
Self-employment is associated with lower wages, reduced retirement savings. Workers are at greater risk of economic shocks and they do not get the protection of many labor standards available to W-2 workers. Then she uses a mixed methods approach, including an index of family policies to address a large variation of state level policies across the US. Her results are that a higher state Family Policy Index score is associated with higher female labor force participation rate. Higher state Family Policy Index score is associated with lower gender gap in labor force participation.
A higher state Family Policy Index score is associated with lower rates of female self-employment, and a higher state Family Policy Index score is associated with lower rates of female solopreneurship. She finds ambiguous quantitative evidence on the impact of state family policies on nonemployer firms for women or solopreneurship. A couple of comments for Kathryn. Again, I enjoyed reading the paper and I’ll send a separate email with suggestions on methodology, particularly ways to exploit the variation in state level family policies a little differently. I also look forward to the qualitative analysis on solopreneurship that you’re starting.
I think it would be interesting to see not just the transition to self-employment, but also those who transition and then came back to traditional employment, because of self-employment not being all that they thought it would be. Interested in what are the differences by worker characteristics, race, education, family size, household income. Self-employment and solopreneurship is not low barrier. There’s high barriers, including financial investments needed and other forms of barriers. How are those manifesting itself and some of the breakdowns by race and family size and so forth. Also, curious about any thoughts you may have on your quantitative analysis and the data you’re using in terms of misclassification issues.
There’s robust evidence that many workers are classified and think of themselves as unemployed, but they really are just, they’re workers, they’re misclassified, so that the employers don’t have to pay some of the taxes and other business taxes they have to pay. Also interested in your thoughts on right-to-work states or states with mandatory arbitration and how that can transition from traditional employment to self-employment, and then just thoughts on states with higher labor standards and high state family policies. For example, here in California, we have a high minimum wage, we value unions, high labor standards, but then you also have this narrative that these policies are not conducive to starting new businesses. Interested in any thoughts you may have on how to square those two. But again, I enjoyed reading the paper.
I’ll move on to paper three, which is Heather’s paper. A quick overview of How Economic Development Incentives Affect Racial and Gender Segregation of Employment and Wages. In her paper, Heather and her co-author explore the impact of EDIs. EDIs are policies like targeted tax breaks, property tax abatements, job creation tax credits, new market brands and so forth on labor market outcomes for women and workers of color. More precisely, she’s asking the question, does the use of EDIs exacerbate occupational segregation and the wage gap by race and gender?
She exploits the variation of EDI across industries and cities to look at the industry share of female and non-white workers and the effect on the male-female wage and white-non-white wage. Some of her results are that EDIs are associated with higher wages within industries, mostly for higher and medium-income wage groups and less so for women. EDIs are associated with larger male-female wage ratios. Industries with higher EDIs are more segregated by race and gender, but not sure which way the causality runs, and she wants to do further segregation to study some of these effects.
A couple of comments for Heather, like before, I really enjoy the paper. I also think that this paper is trying to fill a much-needed research area and I will also provide separate comments on methodology as well. I just wanted to make sure that I understood that when you’re talking about white, you’re talking about non-Hispanic white. I wasn’t clear in the paper whether that was the case. Then, I was also interested in seeing more desegregation by race beyond Black or a non-white category. I would like to read more or get your thoughts more on what you think the mechanisms are that are driving EDIs to impact wages for higher and medium-income wage groups.
Is it that EDIs create more higher-wage jobs? Or could it be that EDIs are leading to gentrification at the expense of lower-income wage groups? For that matter, could gentrification be impacting industry segregation? Interested in any thoughts you have on this. Similarly, I would like to get your thoughts on why women do not benefit as much as men. They said that EDIs are primarily used in industries that are already very heavily segregated male. I know you have an IV to get at this, but I’m wondering if it’s more about the non-randomness in how EDIs are used or the types of EDIs are used or is it that it causes a substitution away from female employment? Again, just interested in any thoughts you have on these mechanisms for your results.
Also, interested in further desegregation of EDIs. I wonder if you can exploit variation in the types of EDIs used within a city to see the effects of those EDI choices on these outcomes. I think, in the end, the important thing that I would say is that EDIs are important tools for cities and states to create jobs, but what really matters is the creation of high quality jobs and jobs with occupational ladders, allowing paths for workers to move from up the wage ladder from lower wage to higher wage jobs, but also improving the quality of lower wage jobs and removing the barriers to better quality jobs, so that it’s not an either or proposition of the types of jobs created. But again, it’s a great paper. I look forward to reading the final draft.
Then finally, to Neil’s paper, Overview of US Labor Force Participation and Unemployment, Structural Change From the Pandemic. Very brief comments. I haven’t had a chance to read it in as much detail, just of the timing of the conference and the paper. He’s looking at where are unequal outcomes and unemployment rates and labor force participation from the pandemic and he looks at pre- and post-pandemic by doing dynamic modeling and forecasting into the pandemic to understand labor market effects of the pandemic. Some of his results are by gender. There’s been a narrowing of the labor force participation rate gap, but this persistent lower labor force participation rates for men and women in the pandemic.
Strong differences of pre-pandemic outcomes by group. Unemployment rates have returned to the pre-pandemic levels by late of 2021. Women in all age groups are more adversely affected than men. Labor force participation rates continue below pre-pandemic for many groups and labor force participation rates by age 55 and plus exited during the pandemic. There was some evidence that they came back to some extent, but now it’s sort of stagnant, the labor force participation rate of these workers. Some very quick comments and then we’ll go to Q&A. I think Neil’s results are consistent with the impact of COVID recession compared to the Great Recession.
The Great Recession was more impactful for men, because of how it happened. Housing and construction jobs were affected and then durable goods that went with housing and manufacturing and all those things made it a male recession, if you will. The pandemic was more female recession, primarily occupational segregation drove a lot of the impact on minorities and women. Just interested in any thoughts Neil has on the 16 to 24 age in terms of the disparities between male and female. There’s been some research showing young men falling behind in terms of education and then translating into some work differences.
But we’ve also seen that there’s been increasing work by younger folks and what that means for educational attainment, just any thoughts that he may have on that. Then also, any thoughts he has on mechanisms for some of the labor force participation challenges that are preventing the return to pre-pandemic levels. Care issues are very important, not just care for children and women with children, but also caring for adult relatives and the prevalence of sandwiched households. Then, housing issues that prevent folks to attain certain jobs. Then, any just precariousness of work.
The pandemic, I think, highlighted for many folks how precarious their work was. What does those preferences for work or changes in those preferences of work, how that may be impacting the labor and force participation rate, challenges that he sees. Then just looking forward to the further analysis on disparities by race, job, and location. With that, maybe we should just give the authors a couple of minutes to respond before taking any questions from the audience. Does that work?
Rajashri Chakrabarti
I can respond really quickly. I basically want to thank you. I think these were really helpful comments and we agree with your interpretation. We do think that what you have said in terms of the mechanism is what is at play, which is why we show that the negative impacts in the higher share Black areas in the household debt market, because of an FFR increase in more distressed areas are not completely explained by negative labor market impacts.
That the balance of it is exactly what you have pointed out, which is really an interaction with pre-existing frictions, inequities that are differential in this market. We see that as a mechanism, the main mechanism. That’s exactly what you have pointed out. I think your comment about comparing with the Great Recession period is a great comment and I think it is definitely a useful exercise, an interesting exercise, to know whether or not this happens every time in a recession or each recessions are different.
Obviously, around that time there were lots of other economic differences that happened. Whether or not they still, we see the same effects or not. That’s not something we have done so far, but it’s an interesting thing that we can definitely do. I’ll keep it there, but I really appreciate your comments and they totally make sense. I was actually chatting with my co-author on the side, and both of us, Maxim and myself, we both thank you a lot for your thoughtful comments.
Enrique Lopezlira
Thanks, Raji.
Kathryn Rudloff
Yeah, I agree. Thank you. That was great feedback and I look forward to, if you can send me some more, I’m still working on this. All input is greatly appreciated at this point. I’ll address your question about the misclassification of self-employed solopreneurs. It is a really big problem and there’s other data sets that are trying to either have more comprehensive when they are relying on individual responses.
Like the NORC is working on the new entrepreneurship in the population study that asked multiple times and really trying to dig down to understand, but that’s why I chose to use the NESD data set, because it is based on administrative data at the individual level. They’re taking, I believe it’s IRS tax data, and kind of matching it up with that demographic information, so that we really could get an understanding of who those nonemployer firms were. It’s a huge data set. I’ve really enjoyed getting to dive into it and I welcome the opportunity to talk about it with other researchers interested in leveraging that sort of access.
Enrique Lopezlira
Thanks, Kathryn.
Heather Stephens
Okay, I’ll go ahead and go. Thank you so much Enrique for your comments. This is a pretty new paper, so it’s very useful at this stage to get some good feedback. Yes, when we say white, it is non-Hispanic white, we made sure that that was the case. I also saw an audience person had asked if the more incentives is associated with lower shares of minority employment, and yes, that is the case. Potentially, the reason is political. Those particular results are not the regression analysis, those are just correlations.
In our regression results we really are trying to control for this kind of reverse causation where majority white industries might have more political power potentially to get more incentives or things like that. But overall, there is lower shares of minority employment and incentivize industries. Then, there was also another audience question about which industry types are generally incentivized. There’s lots of variation across space in terms of which industries are incentivized, but with our statistical methods, we’re actually controlling for industry and year trends.
What our results basically suggest is that incentivizing specific industries or specific industries within a city can be interpreted as increasing inequality in that industry, in that city. It’s after controlling for these trends overall. Again, you can look at what the big ones are, a lot of manufacturing industries and things like that, but we should be controlling for that and we still are seeing these increasing inequality and increasing disparities in terms of wages. Thank you again.
Enrique Lopezlira
Thanks, Heather.
Merissa Piazza
I’m going to jump in. Thank you to everybody for such a robust discussion. On behalf of the organizing committee, we want to thank you for joining us. I want to acknowledge all of our participants for their time and insights shared today and the informative discussion. In a moment, you’ll see a link to our daily survey. Please provide feedback on the session as you see fit. There was a question and answer in the chat on when information will be posted. Please consult that and thank you again. Have a great day.